UNIVERSITY OF CALIFORNIA RIVERSIDE Modeling the Size Frequency Distributions of the Trilobite Aulacopleura koninckii and its Implications for Understanding Trilobite Biology and Preservation Potential A Thesis submitted in partial satisfaction of the requirements for the degree of Master of Science in Geological Sciences by Rachel Lynn Kolenko March 2016 Thesis Committee: Dr. Nigel C. Hughes, Chairperson Dr. Gareth Funning Dr. Richard A. Minnich Copyright by Rachel Lynn Kolenko 2016 The Thesis of Rachel Lynn Kolenko is approved: Committee Chairperson Acknowledgements Most importantly, I would like to thank my advisor, Dr. Nigel Hughes, for both his support and insight during my time as a graduate student. Thank you for helping me grow as a paleontologist and giving me the opportunity to investigate my research to the fullest extent. I would like to acknowledge my other committee members, Dr. Gareth Funning and Dr. Richard Minnich for their input during the thesis writing process. Many thanks also goes out to the friends I have made during my time at the University of California, Riverside for the unforgettable memories we created. I would like to recognize the University of California, Riverside especially the Department of Earth Sciences for providing me with the means necessary to pursue my graduate education. And finally, I would like to thank my family. Without their perpetual support I would not be where I am today. iv ABSTRACT OF THE THESIS Modeling the Size Frequency Distributions of the Trilobite Aulacopleura koninckii and its Implications for Understanding Trilobite Biology and Preservation Potential by Rachel Lynn Kolenko Master of Science, Graduate Program in Geological Sciences University of California, Riverside, March 2016 Dr. Nigel C. Hughes, Chairperson Trilobites were a diverse group of Paleozoic marine arthropods, a group whose growth is characterized by exoskeletal ecdysis, or the molting of the outer cuticle. If all sclerites from all growth stages were preserved in a fossil assemblage, the distribution would be expected to be strongly right skewed because all individuals must have pass through smaller molt stages. While not all juveniles survive to large size, the overwhelming majority of observed trilobite size frequency distributions have normal distributions. This study investigates this disparity using a modified version of the method used by Hartnoll and Bryant (1990) to model crab size frequency distributions v based on the parameters of instar duration, mortality rate, sizer per instar, and number of instars. Here I apply this method to the trilobite Aulacopleura koninckii, a Silurian species whose growth is among the best known for any fossil, and for which size-specific assemblages of articulated individuals are recorded. This required a combination of parameters known for A. koninckii, in combination with other estimates based on living crab biology. Observed growth parameters from A. koninckii suggest that this trilobite underwent up to 33 post-potaspid instars separated by 32 post-potaspid molts. A range of low mortality rates, 5%, 10%, 15% at the first meraspid were assumed based on evidence that this taxon occupied a low predation environment. Assuming a constant recruitment and crab-based estimates of systematic changes in inter-molt duration, I was able to model predicted distributions of A. koninckii sizes that matched the largest individuals observed. Estimated life spans of A. koninckii according to these parameters ranged between 1 and 20 years. Although fitting the observed size range these distributions predicted far larger numbers of smaller specimens than larger ones, but this was not observed in the fossil record. To account for the dearth of smaller individuals, I explored the effect of selective preservation. Simulations suggest that the size frequency distributions observed for A. koninckii could be the outcome of either preservation bias against smaller trilobites or events that killed young populations. vi Table of Contents Title Page ……………….. i Signature Page ……………….. iii Acknowledgements ……………….. iv Abstract ……………….. v Introduction ……………….. 1 Background ……………….. 1 Materials and Methods ……………….. 10 Results ……………….. 15 Discussion ……………….. 38 Conclusion ……………….. 43 References ……………….. 45 Appendix A ………………..47 vii List of Figures Figure 1 ……………….. 4 Figure 2 ……………….. 6 Figure 3 ………………..17 Figure 4 ……………….. 18 Figure 5 ……………….. 19 Figure 6 ……………….. 22 Figure 7 ……………….. 23 Figure 8 ……………….. 24 Figure 9 ……………….. 27 Figure 10 ……………….. 28 Figure 11 ……………….. 29 Figure 12 ………………..31 Figure 13 ……………….. 32 Figure 14 ……………….. 34 Figure 15 ……………….. 37 viii List of Tables Table 1 ……………….. 25 Table 2 ……………….. 25 Table 3 ………………..26 ix Introduction Size frequency distribution plots are often used to report the occurrence of organisms in a single population or assemblage belonging to a single species. With regard to fossil data, a size frequency distribution can convey both biological and preservation data about a specific taxon, reflect variations among species that may indicate different life histories, and thus reveal aspects of paleoecological habits. In trilobites, size frequency distributions are complicated by their molting habit. Rarely is it known whether fossil material represents fresh carcasses killed during burial, individuals dead prior to burial, or exuviae: the remains of the exoskeleton after molting. This study uses simulated models of arthropod growth, based on those of Hartnoll and Bryant, 1990, to explore different aspects of trilobite life history, constrained by our best knowledge of trilobite biology, and the use of reasonable comparison with living arthropods. Using these constraints I produced a range of size frequency distributions that are within plausible bounds for the trilobite, Aulacopleura koninckii, a species whose ontogeny and occurrence is well documented. Background To explore the possible controls of ancient size frequency distributions, it is necessary to constrain the varied inputs that could account for such distributions. This requires focusing on those species whose growth is well known. Many living arthropods grow according to Dyar’s rule (1890), which is the observation that per-molt size increases occur at a constant growth increment (Fusco et al. 2012). Such a pattern of 1 growth is also evident among trilobites, with values of per-molt growth increments varying between 1.03 and 2.21 in the meraspid growth stages (Fusco et al. 2012). The growth increment of A. koninckii is particularly well constrained, due to the unusual site specific preservation of hundreds of articulated specimens spanning almost the full range of post-protaspid ontogeny; for total body length it is estimated as 1.12 (Hughes et al. 2014). During the meraspid period, in which segments accreted progressively in the thorax, it is possible to precisely determine individuals between instar (molt phase) growth increment, which are remarkably constant (Fusco et al. 2004). Furthermore, variance in size within instars is also notably constant (Fusco et al. 2004). The constant meraspid molt increment provides a justifiable basis of projecting into the following holaspid stage in order to estimate the number of molts required to account for the largest observed individual. The number was 32, with the longest individuals being about 28 millimeters in length. Previous studies conducted by Hartnoll (1978), Sheldon (1988), and Hartnoll and Bryant (1990) explored various size distribution characteristics of fossilized trilobite and modern crab populations and accumulations. These were based on both observed growth and tempo of molting increments, i.e. the time interval between molts, under varied sea water temperature and salinities. Hartnoll (1978) also modeled the cumulative, standing size frequency distribution of a crab population based on a model that considered projected growth increments and molt tempo, at assumed constant recruitment and mortality. Surprisingly, Hartnoll (1978) showed from observed growth parameters that the size distributions of living crabs are predicted to be normal with high infant mortality. 2 As the rate of molting decreases, the numbers of individuals accumulate at given larger sizes or instar numbers. Hartnoll and Bryant (1990) built on Hartnoll’s previous work by modeling the frequency distributions, both instar and size, of decapod populations. They used the parameters instar duration, mortality, and average carapace width per instar to produce frequency distributions for crab species Cancer anthonyi. They modeled 4 different types of distributions; the population of live crabs, the number of corpses, the exuviae produced, and a combination of exuviae and corpses. Hartnoll and Bryant (1990) found that the frequency distributions of live populations of crabs differed markedly from the distributions of their corpses, exuviae, and the mixture of both corpses and exuviae (Figure 1). 3 Figure 1: Frequency distributions by instar (left column) and size (right column) from Hartnoll and Bryant, 1990 As determined previously, distributions of live individuals had a normal distribution due to each subsequent intermolt period in an organism’s life cycle being longer in duration than the previous intermolt. This caused a population in equilibrium to have a normal distribution as individuals from multiple cohorts/ generations pooled at the larger instars. The three remaining distributions have a right skew, reflecting the accumulated remnants of the crab populations,